摘要
经典多元回归分析不能实时跟踪响应变量的动态变化,而且样本中少量病态数据的出现影响拟合效果。针对该问题,运用灰色组合预测方法剔除了自变量观察数据中的噪声污染,对传统的多元线性回归分析方法进行了改进,建立了灰色组合多元线性回归分析模型。最后,通过实例说明模型具有较高的预测精度。
Classical multiple linear regression analysis cannot track the response variables dynamic change in time and the fitting results would be impacted for a small amount of sick data. This paper applies grey combination forecast method to reject noise pollution in the observation data of independ- ent variable, improves the traditional method of multiple linear regressions, and sets up a grey combination multiple linear regressions model. Finally, the more predictable precision examples show this improvement model has
出处
《重庆理工大学学报（自然科学）》
CAS
2012年第8期113-116,共4页
Journal of Chongqing University of Technology：Natural Science
基金
国家社会科学基金西部项目(11XTJ001)
关键词
灰色模型
组合预测
多元线性回归
grey predication model
combination forecast
multiple linear regressions
